Testing the Prediction Studio APIs

3 Tasks

10 mins

Pega Customer Decision Hub '25

Visible to: All users Applies to: Pega Customer Decision Hub '25

Beginner

English

Scenario

The Prediction Studio APIs facilitate the monitoring and management of predictions remotely, using the scripting language of your choice. They offer endpoints for various operations, such as getting a list of Predictions and their attributes, introducing a challenge model, and more. These operations are categorized under Data export, Models, Reporting, Predictions, and Prediction Studio. As a System Admin, you generate the client credentials that provide access to the APIs, and then inspect the API responses.

Use the following credentials to log in to the exercise system:

Role	User name	Password
System Admin	SystemAdmin	rules

Your assignment consists of the following tasks:

Task 1: Generate the ADM export artifacts

Generate the dataflows to export the ADM model and predictor snapshots.

Task 2: Generate a Client ID and a Client Secret

Generate a Client ID and a Client Secret for access to the Prediction Studio APIs.

Task 3: Inspect the API responses

Use REST calls to try out the Prediction Studio APIs.

 

You must initiate your own Pega instance to complete this Challenge.

Initialization may take up to 5 minutes so please be patient.

• Log In

Challenge Walkthrough

Detailed Tasks

1 Generate the ADM export artifacts

1. Log in as a System Architect:
   1. In the User name field, enter SystemAdmin.
   2. In the Password field, enter rules.
2. In the header, click DEV STUDIO, and then switch to PREDICTION STUDIO.
3. In the navigation pane of Prediction Studio, click Settings > Prediction Studio settings.
4. Scroll down to the Monitoring database export section, and then click Configure to open the Export monitoring database dialog box.
5. In the Export monitoring database dialog box, click Submit to generate the export artifacts.
6. Close the Export monitoring database dialog box.
7. In the header, click PREDICTION STUDIO, and then switch to DEV STUDIO.

2 Generate a Client ID and a Client Secret

1. In the header of Dev Studio, click Create > Security > OAuth 2.0 Client Registration to create a new OAuth 2.0 Client Registration:
2. In the OAuth 2.0 Client Registration short description field, enter Prediction APIs.
3. In the Client Name field, enter PegaClient.
4. Click Create and open.
5. Configure the Client Registration:
   1. In the Supported grant types section, select Client credentials, and then, in the Operator ID field, enter or select DataScientist.
   2. In the Client Credentials section, click View & download, and then click Download credentials.
   3. Securely store the PegaClient_ClientCredentials text file that contains the Client ID and the Client Secret for later reference.
6. Close the dialog boxes, and then click Save to complete the OAuth 2.0 Client Registration.

3 Inspect the API responses

1. In the header of Dev Studio, click Application: CDH > Channels and interfaces.
2. In the Current channel interfaces section, click API.
3. In the Service package field, select Prediction APIs, and then click Authorize.
4. In the CLIENTCREDENTIALS (OAUTH2, CLIENTCREDENTIALS) section, enter the Client ID and the Client Secret from the PegaClient_ClientCredentials text file that you saved in the previous task, and then click Authorize.
   
5. Confirm that the client is authorized, and then click Close.
6. In the Models section, click the GET /V1/models API to list the models that are part of the application.
   
7. Click Try it out, and then click Execute.
8. Review the content of the server response body, and take note of the modelId of the Web Click Through Rate GB Customer adaptive model.
   
9. In the Models section, click the GET /v1/models/{modelID} API to get a description of the model, and then click Try it out.
   
10. In the modelID field, enter DATA-DECISION-REQUEST-CUSTOMER-CDH!WEB_CLICK_THROUGH_RATE_GB_CUSTOMER to get a description of the Web Click Through Rate Customer adaptive model, and then click Execute.
11. In the server response body, review the modeling technique, the target and alternative labels, and the predictors of the adaptive model.
12. In the Data export section, click the POST /v1/datamart/export API to trigger the export of the datamart.
    
13. On the enablement landing page, click Application Switcher > Filebrowser.
14. Log in to the repository:
    1. Username: pega-filerepo
    2. Password: pega-filerepo
15. Open the datamart folder and inspect the exported model snapshots and predictor snapshots files.

---

This Challenge is to practice what you learned in the following Module:

• Monitoring adaptive models v9

---

Available in the following mission:

• AI for 1:1 Customer Engagement v6